Posts made by bardicverse

@Wayne-Workman Thanks for that. Yes, running Ubuntu 12. In a catch 22, would need to wipe this server, but running osticket for tech ticket support. will have to bring that down to install newer os. But also need to get these computers imaged before the new school year starts. Going to have to bite the bullet I guess. I do have a few older pcs that might be able to handle the task too. Decisions, decisions.

Trying to download kernel 4.1.2 so my Optiplex 5040s will work properly. When I go to the kernel update page, there’s a list of several kernels from 2015, but no download links. Only the older ones from 2014 and earlier have download links. Anyone have a working link to the 4.1.2 kernel download? Need to get these systems imaged.

Thanks

So, I found a very odd but simple solution. Apparently hitting the Enter key on the keyboard advanced it to the next screen, and started uploading. Odd, but it worked.

Had to reinstall ubuntu and fog after some major network issues. The system worked prior to the issues without a problem.
System is Ubuntu 10.04 , Fog 1.2.0, on an Optiplex 960 desktop.

On fog install, I left the mysql database passwords blank, didn’t change them yet.
Updated schema, all good to go, linked my external USB hard drive where the images are kept and mounted it. Created new image settings in fog, linked to the proper folders on the external hard drive.
Registered host, sent a command to upload an image to one of the image folders I linked prior, gets through to the “Using Hard Disk” message, just before going to next screen for the upload progress, and freezes there.

I am certain I missed something fairly simple, like changing the database username and passwords, but the config file in fog/commons is missing in 1.2.0. Any ideas where that’s moved to? Also, anything else I’m missing to make the upload work?

So I’ve scoured the forums here regarding cloning Acer netbooks, and couldn’t find an answer.

I am able to get an image from a netbook uploaded, no problems. When I go to clone based on that image, everything connects through, and it goes to check the hd disk status, and reboots, marking the task as complete, and obviously is not the case. The screen flashes by in a millisecond and can’t grab any sort of information or error in time.

I really would like to use fog with these netbooks as we have 20 of them for our school. Fog works great everywhere else, from custom built rigs that still exist here to Dell Optiplex 745 and up.

Any thoughts on what I can try/check to make these computers actually image properly?

So, if I get this error when preparing to upload, it’s going to be a problem on the clones too? If that’s the case, what’s the easiest way to fix the host machine, so that this error doesn’t pop up anymore?

[quote=“Tom Elliott, post: 29273, member: 7271”]fixparts is also a part of the init files.[/quote]

So, it can be run from the fog prompt? Is this only going to happen with uploading an image, or will it happen for the cloning process to? Getting ready to build images for the new school year.

yup tried that too, doesnt fix the issue. The problem is the workflow - client computer set to upload a new image - folder is created under /images/dev, then is transported to its final spot /images/(image name) . This normally works flawlessly. Yet, it seems that because “root” is the owner of the temporary folder, it wont allow “fog” to copy it and then remove it from its temporary spot. Since this process in fog is consecutive and unattended, there’s no time to pause and chmod or chown the folder.

Basically I need to make “fog” be the users who creates the temporary image folder. Yet, right now, its “root” who has that honor. I need to figure out where I can assign “fog” for the folder creation process.

I’ve scoured the internet and have tried many solutions to my “unable to move /images/dev/(macaddress) to /images/(image folder)” issue, to no avail.
After poking and prodding, I stumbled across something. The new image folder in /images/dev has a different owner when the folder is created. These temporary folders are created by “root”. When I chowned -R the folder to the user “fog”, the “unable to move” issue went away.

Yet, every time I upload a new image, “root” owns the file in in /images/dev, which causes the issue all over again.

How do I assign the default creator to “fog”? I’m not sure if this is in one of the fog installation configs, or if its a general linux thing. Fog version 0.32 and Ubuntu 10.04 LTS.

No, I have the working kernel that needs to be swapped out, but is the bz_image file something separate? Or is that part of the kernel?

Yup, I need to register hosts that are Dell Optiplex 760 models. So, what would I be replacing the bzImage file with?

btw using fog 0.32 and ubuntu 10.04

So our 755 lab imaged well with Fog, no issues. Went to prepare our 760 lab to PXE boot, and found it freezes after it acquires the data from the fog server. Some research told me that an older kernel worked on this model for other people, but the newer kernel wouldn’t. I found the working kernel file, now I need to figure out how to swap it out on the fog server.

So, where do I need to put this kernel for it to send out to the PXE booting machines? I figure I’ll use this kernel to get the imaging process done on this lab, and then shut off PXE boot on them afterwards, and go back to the newer kernel.

Seems weird your 745 is doing that. I have the same setup (10.04 and fog 0.32) and got my 745s re-imaged without any issues. I do get it on my 990s.

Another user suggested that the hard drive might be set up for channel 1 as opposed to 0 in the bios/physical cabling. Food for thought anyhow. Don’t eat the sea poop, albeit enticing. =p

dumb question - can you ping the fog server’s ip address from one of your client machines?

not sure what version of syslinux. To be honest, I can always just use the one time boot menu when i need to reimage, etc. Right now I’m just trying to roll out these images asap.

I found a similar problem, but my solution was too vague, because I made several changes. Of those, I did the following:
-Switched from Ubuntu 12 to 10.04
-On my DHCP server, set port 66 to my IP address, not the server name
-modified Ubuntu with a proxydhcp

One or more of those things worked for me, on a windows client.

actually, that would probably be the case. Dell hard drives come partitioned with a recovery partition, or disk 0, and the next partition is the main body partition, which would probably be disk 1. I’ll have to test that theory out later. currently uploading 2 new images to the server.

I think in future practice, I’m going to wipe the dell recovery partition outright, so I can use the resizable image option. But since we’re already underway with the summer workload, it’ll have to wait.

makes sense. Ah well, I’ll be poking and prodding on it over the summer once I got these other labs reimaged. Will post back if I come up with a solution.

thanks for that. and yup I got what you meant. I thought it might be a disparity from 32 bit to 64. Are your 390s 32 bit by chance?

My server is 32 bit and the 990s are 64 bit. Though, trying the chainboot fix only caused my computer to not restart, just reloop the fog screen itself over and over. So, I’m not sure what’s going on. Have you tried a 64 bit client yet?